Full text of DOI:10.1385/BTER:74:1:55

©Copyright 2000 by Humana Press Inc.
All rights of any nature, whatsoever, reserved.
0163–4984/00/7401–0055 $14.00
Selenium Regulates Expression
in Rat Liver of Genes for Proteins
Involved in Iron Metabolism
MERRILL J. CHRISTENSEN,* CARI A. OLSEN,
DAVID V. HANSEN, AND BLAKE C. BALLIF
Department of Food Science and Nutrition,
Brigham Young University, Provo, UT 84602
Received June 18, 1999; Accepted July 15, 1999
ABSTRACT
Suppression subtractive hybridization analysis in our labora-
tory recently revealed that transferrin mRNA may be elevated in Se-
deficient rat liver. In this work, we compared expression in rat liver
of genes for transferrin, transferrin receptor, ferritin light and heavy
chains, and iron-regulatory proteins 1 and 2 in Se adequacy and de-
ficiency. Weanling male Sprague–Dawley rats were fed Torula yeast
diets supplemented with 0 or 0.15 µg Se/kg diet as sodium selenite
for 15 wk. Activity of cellular glutathione peroxidase was virtually
abolished in Se-deficient rat liver, whereas activity of glutathione
S-transferase was 43% higher than in Se-adequate liver. There were
no differences in hematocrit, hemoglobin, or liver iron content. To
examine differential gene expression, we used a multiplex relative
reverse transcriptase–polymerase chain reaction method. Three of
the six genes examined showed modest but consistent upregulation
in Se deficiency. Transferrin mRNA was 30% more abundant in Se-
deficient than in Se-adequate liver. For the transferrin receptor, the
difference was 32%, and for iron regulatory protein 1, it was 63%.
No consistent differences were observed for iron regulatory protein
2 or for ferritin light or heavy chain. These findings suggest a pos-
sible role for dietary Se in moderating iron metabolism.
Index Entries: Selenium; gene expression; rat; transferrin; trans-
ferrin receptor; IRP 1; RT-PCR.
*Author to whom all correspondence and reprint requests should be addressed.
Biological Trace Element Research
55
Vol. 74, 2000

56
Christensen et al.
INTRODUCTION
The essential trace element selenium (Se) functions biochemically as a
component of selenoproteins, in which Se is incorporated into the primary
structure of the protein as selenocysteine (1). Selenoproteins of known
function include members of the glutathione peroxidase (GPX) (EC
1.11.1.9) family (2–6), which function as antioxidants differing in substrate
specificity and tissue distribution. Iodothyronine 5Ј-monodeiodinase (EC
3.8.1.4) Types I (7), II (8), and III (9) are selenoproteins with essential func-
tion in thyroid hormone metabolism. Thioredoxin reductase, another
selenoprotein (10), is a flavin adenine dinucleotide (FAD)-containing
enzyme that forms a redox system with its substrate, thioredoxin.
Although some of the biological effects of Se can be attributed to the
actions of previously characterized selenoproteins, many cannot. The ef-
fects of Se on hepatic heme metabolism (11), glutathione metabolism (12),
cytochrome P-450 induction (13), activity of drug-metabolizing enzymes
(14), protection against heavy-metal toxicity (15), essentiality in immune
function (16), and its cancer chemopreventive effects (17) have all been
shown to be independent of GPX activity. The biochemical and molecu-
lar mechanisms by which Se exerts these effects are, in many cases, un-
certain. To explain Se’s previously observed biological effects and to define
new biochemical or molecular roles for this element, we have sought to
identify genes whose expression is regulated by dietary intake of Se.
Techniques previously used to characterize differentially expressed
genes include differential screening (18), subtractive library construction
and hybridization (19), differential display (20), and RNA fingerprinting
(21). In this work, we employed the more recently developed technique
of suppression subtractive hybridization (SSH) (22) to identify Se-regulated
genes in rat liver. This technique combines traditional subtraction with
selective amplification of differentially expressed cDNAs by polymerase
chain reaction (PCR). By this process, a cDNA for transferrin was shown
to be present at different levels in livers of rats fed diets adequate and
deficient in Se. This prompted our examination of the effect of Se defi-
ciency on expression of genes for additional proteins involved in iron
(Fe) metabolism.
MATERIALS AND METHODS
Animals
Weanling male Sprague–Dawley rats (Harlan Sprague–Dawley, Indi-
anapolis, IN) were housed in pairs in hanging wire cages in a tempera-
ture- and light-controlled room (12 h light; 12 h dark). Animals were
given free access to food and tap water. All procedures related to animal
use were in accordance with the Guide for the Care and Use of Laboratory
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
57
Table 1
Composition of Torula Yeast Basal Diet
1Dyetrose (Dyets, Bethlehem, PA).
2Solka-Floc, 200 FCC (FS&D, St. Louis, MO).
3Mineral mix was formulated as described (24) for AIN-93G-MX with the fol-
lowing modifications: (a) sodium selenate, anhydrous was omitted; (b) potassium
phosphate, monobasic was omitted; (c) potassium citrate, tripotassium, monohydrate
was increased from 70.78 to 226.50 g/kg mix; (d) powdered sucrose was increased
from 221.026 to 261.306 g/kg mix. See text for discussion of dietary P content.
4Vitamin mix was formulated as described (24) for AIN-93G-VX with the fol-
lowing modification: Vitamin E (all-rac-α-tocopheryl acetate) (500 IU/g) was in-
creased from 15.00 to 20.00 g/kg mix.
Animals (23) and were approved by the Brigham Young University Ani-
mal Care and Use Committee.
Diets
The composition of the basal Se-deficient diet is shown in Table 1.
This mixture was formulated as described (24) for the AIN-93G diet, with
the following modifications:
1. Torula yeast (Lake States Division, Rhinelander Paper Co.,
Inc., Rhinelander, WI) at 30% was substituted for casein at
20% to provide a low-Se protein source.
Biological Trace Element Research
Vol. 74, 2000

58
Christensen et al.
2. DL-Methionine was substituted for L-cystine, as methionine
is the limiting amino acid in Torula yeast (25).
3. Nonspecific sources of carbohydrate (cornstarch, dextrinized
cornstarch, sucrose) were each reduced proportionately to
compensate for the increase (from 20% to 30%) in the pro-
tein source.
4. Se was omitted from the mineral mix in the basal Se-deficient
diet; in the Se-adequate diet, Se was added as sodium sele-
nite to provide 0.15 mg Se/kg diet.
5. Vitamin E was included in the vitamin mix to provide 100
(rather than 75) IU/kg diet.
6. Potassium phosphate was omitted from the mineral mix to
limit the amount of phosphorus to that recommended for the
AIN-93G diet; potassium citrate, tripotassium, monohy-
drate (36.16% K) was increased to provide the recommended
amount of potassium.
The AIN 1993 formulation provides a total of 5 g Ca and 3 g P/kg
diet, for a Ca : P molar ratio of approximately 1.3. Certificates of analysis
provided by various suppliers indicate that Torula yeast is an abundant
source of P, containing 1.5–2.0% of this element. The same certificates of
analysis report that Torula yeast contains Ca at a level ranging from 0.015
to 0.95%. Thus, Torula yeast provided at 30% would, by itself, supply
4.5–6.0 g P and up to 2.85 g Ca/kg diet. We included Ca in the mineral
mix as specified by AIN-93 to provide 5 g Ca/kg diet, but omitted P. This
combination would result in a diet providing a total of 5–7.85 g Ca and
4.5–6.0 g P for a Ca : P molar ratio of 0.64–1.35. This is a considerable
improvement over previous formulations of Se-deficient diets, in which
the AIN-76A mineral mix providing 4 g P/kg diet was used. In those
diets, assuming the same contributions of Ca and P from Torula yeast,
the calculated Ca : P molar ratio would have ranged from 0.39 to 0.71.
Animal Processing
Rats were fed in groups of 20 for 15 wk. Animals were weighed
weekly. At the end of the feeding period, rats were killed by decapitation.
Whole blood was collected and a small aliquot was used for hematocrit
measurements. The remainder was stored at −20°C. A small piece of each
liver was homogenized in 3 volumes of ice-cold 0.25 mol/L sucrose.
Homogenates were centrifuged at 105,000g for 60 min and the resulting
supernates were stored at −80°C until assayed for enzyme activity. The
remainder of the livers were flash frozen in liquid nitrogen and stored
frozen at −80°C until needed for RNA isolation or total Fe determinations.
Iron Status Indicators
Fresh whole blood was used for determination of hematocrit (26).
Frozen samples were thawed and hemoglobin concentrations were
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
59
determined using a commercially available diagnostic kit (Sigma, St.
Louis, MO). Liver Fe concentrations were measured by the method of
Torrance and Bothwell (27).
Assay of Glutathione Peroxidase
and Glutathione S-transferase
Activity of cellular glutathione peroxidase (GPX1) was assayed as
reported previously (28) according to the coupled method of Lawrence
and Burk (29), using 2 mmol/L of glutathione (GSH), and 0.25 mmol/L
of H2O2 as substrate. One unit of enzyme activity was defined as 1 µmol
NADPH oxidized per minute. Activity of glutathione S-transferase
(EC 2.5.1.18) (GST) in liver cytosol was assayed as reported previously
(30) according to the method of Habig et al. (31) using 1-chloro-2,4-
dinitrobenzene as the substrate. One unit of enzyme activity was defined
as 1 µmol product formed per minute. Protein concentration of cytosols
was determined by the method of Lowry et al. (32).
All chemicals used in these and other assays were purchased from
Sigma Chemical Co. (St. Louis, MO). Substrates for enzyme assays (1-
chloro-2,4-dinitrobenzene, GSH, NADPH) were supplied at purities rang-
ing from 95% to 100%. All other chemicals were reagent grade or purer.
mRNA Isolation
Total RNA was isolated from tissue samples using Trizol® reagent
(Life Technologies, Gaithersburg, MD, USA). The yield of total RNA from
each sample was determined by optical density measurement at 260 nm.
Equal aliquots of RNA from each of three individual livers within a
dietary group were combined to form Se-adequate and Se-deficient rat
liver RNA pools. RNA pools were treated with DNase (RQ1 RNase-free
DNase, Promega Corp., Madison, WI, USA) in the presence of a RNase
inhibitor (Promega) according to the manufacturer’s instructions. Follow-
ing DNase treatment, RNA pools were further purified using RNeasy®
columns (Qiagen Inc., Valencia, CA, USA) according to manufacturer’s
directions. The concentration of RNA was again determined spectropho-
tometrically and integrity was verified by formaldehyde agarose gel elec-
trophoresis with visualization by ethidium bromide (EtBr) staining. From
each pool, mRNA was isolated using Oligotex™ resin (Qiagen).
Identification of Differentially Expressed Genes
Genes expressed differentially in Se adequacy and deficiency were
identified by SSH (22) using a commercially available kit (PCR-Select™
cDNA Subtraction Kit, CLONTECH Laboratories, Inc., Palo Alto, CA,
USA) with minor modifications in hybridization conditions. The cDNAs
included in the primary hybridization were ethanol precipitated together,
then dissolved in 1.5 µL of hybridization buffer (50 mmol/L HEPES,
Biological Trace Element Research
Vol. 74, 2000

60
Christensen et al.
0.02 mmol/L EDTA, 0.5 mol/L NaCl, 10% PEG 8000) as described by
Diatchenko et al. (22). The secondary hybridization was begun by addi-
tion of 1 µl of denatured driver cDNA and 0.4 µl of the proprietary
CLONTECH hybridization buffer included in the kit.
Subtracted Se-adequate and subtracted Se-deficient cDNAs were
PCR amplified and a fraction of each was cloned into the pCR® II vector
(Invitrogen Corp., Carlsbad, CA, USA). Miniprep plasmid DNA was pre-
pared from each clone (QIAprep® Spin Miniprep Kit, Qiagen) and dot
blotted in duplicate onto two nylon membranes (MagnaGraph, Micron
Separations Inc., Westborough, MA, USA). The remaining PCR-amplified
subtracted cDNAs, both Se-adequate and Se-deficient, were radiolabeled
as probes (DECAprime™ II Random Priming DNA Labeling Kit, Ambion,
Inc., Austin, TX, USA). The membranes were incubated with either the
Se-adequate or the Se-deficient probe overnight and washed according to
the membrane manufacturer’s instructions. Clones whose signals were
markedly different on duplicate membranes were judged to be differen-
tially expressed. These cDNAs were sequenced (T7 Sequenase™ Quick
Denature Plasmid Sequencing Kit, Amersham Pharmacia Biotech, Inc.,
Piscataway, NJ, USA) and their sequences submitted for similarity
searches in GenBank using the NCBI BLAST programs.
Confirmation of Differential Gene Expression
A relative multiplex reverse transcriptase–polymerase chain reaction
(RT-PCR) assay was used to confirm differential gene expression. In this
protocol, 5 µg total RNA from each DNase-treated pool were reverse
transcribed in 50-µL reactions using Superscript™ II reverse transcriptase
(Life Technologies) and 1 µmol/L random hexamers as primer. Follow-
ing reverse transcription, PCR amplification was performed by combin-
ing 36 µL PCR-grade water with 1 µL of 10 mmol/L dNTPs (Promega),
1 µL each of 50 µmol/L forward and reverse primers (see Table 2), 5 µL
cDNA from the RT reaction, 5 µL of 10X PCR buffer, and 1 µL of 50X
Advantage®2 Polymerase Mix (both from CLONTECH). Following thor-
ough mixing of all components, each reaction mixture was then dis-
pensed into six aliquots of 8 µL in 0.2-mL thin walled PCR tubes and
overlaid with oil. Temperature cycling was performed in a DeltaCycler
II™ System (Ericomp, San Diego, CA, USA). Cycling parameters were as
specified in the Advantage®2 user manual: 1-min denaturation at 95°C
followed by cycles of 95°C for 30 s, annealing and extension at 68°C for
1 min, and at the conclusion of cycling, an additional extension at 68°C
for 1 min, and a final 5-min incubation at 72°C.
Thermal cycling was initiated for one aliquot of each sample. After
one cycle, the second aliquots were added to the thermal cycler; after two
cycles, the third aliquots were added, and so on until all six aliquots had
been added. Thus, each sample was amplified over a range of cycles.
With 10 cycles remaining, forward and reverse primers for 18S rRNA
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
61
Table 2
Sequences of PCR Primers
cDNA were added to each aliquot to a final concentration of 400 nmol/L.
At the end of the cycling protocol, each aliquot was electrophoresed
through EtBr–agarose gels. Bands in each gel were visualized and their
intensities quantitated using an AlphaImager™ digital imaging system
(Alpha Innotech Corporation, San Leandro, CA, USA). Intensities of the
bands of interest, normalized for 18S band intensities, were plotted as a
function of cycle number. Regression equations fitted to the curves were
used to calculate the number of cycles necessary for the normalized flu-
orescence from each sample to reach a detection threshold. Differences
between samples in the number of cycles required to reach that thresh-
old reflected differences in the original abundances of those species.
Statistical Analysis
Data were subjected to Student’s t-test for analysis of differences
between dietary groups in Fe status indicators and enzyme activities.
Biological Trace Element Research
Vol. 74, 2000

62
Christensen et al.
A p-value of less than 0.05 was accepted as the level of statistical
significance.
RESULTS
Growth, Enzyme Activities,
and Fe Status Indicators
At the end of the 15-wk feeding period, body weights (mean SEM)
for the Se-deficient and Se-adequate groups were 382 7 and 408 8 g,
respectively. This difference was statistically significant (p < 0.001). Activ-
ity of cellular GPX1 was 535 27.5 mU/mg protein in Se-adequate rat
liver, but only 1.01 1.50 mU/mg protein in Se-deficient liver. The activ-
ity of GST in Se-adequate and Se-deficient rat liver was 229 13.8 and
328 18.4 mU/mg protein, respectively. This difference was significant
(p < 0.01). The virtual abolishment of GPX1 activity and the higher
GST activity in livers of rats fed the Se-deficient diet confirmed their Se-
deficient status. There were no statistically significant differences be-
tween dietary groups in hematocrit, hemoglobin, or total liver Fe content.
Dot-Blot Screening
Figure 1 shows the results of a dot-blot hybridization performed as
described earlier. The circled signals (E3 and H2) are from cDNAs that
appear to be present at higher levels in samples from Se-deficient rat
liver than from Se-adequate livers. When these cDNAs were sequenced
and submitted to GenBank, BLAST homology searches showed a 92%
nucleotide identity for both with rat transferrin (Tf) mRNA. The discov-
ery that Tf mRNA levels may be regulated in part by dietary Se intake
prompted us to examine the expression of genes for other proteins
involved in Fe metabolism in Se-deficient and Se-adequate rat liver.
Confirmation of Differential Expression
Validation of the method of multiplex relative RT-PCR for demon-
stration of differences in gene expression was provided by examination
of gene expression for GPX1. It is well established that mRNA levels for
this protein are markedly lower in Se-deficient rat liver than in Se-
adequate liver (33). Figure 2 shows the EtBr–agarose gel of GPX1 cDNA
amplified from Se-adequate and Se-deficient rat liver RNA samples over
a range of cycles and the coamplified cDNA for 18S rRNA. The best-fit
exponential curves for normalized signal intensities were plotted for both
Se-adequate and Se-deficient RNA and the equations are shown. Assum-
ing that the amount of product doubles with each cycle of PCR, the dif-
ference in original abundance is calculated as 2n where n is the difference
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
63
Fig. 1. Dot blots of cDNAs suggested by SSH analysis to be differentially
expressed in rat liver in Se adequacy and deficiency. cDNAs upregulated in Se
adequacy and cDNAs upregulated in Se deficiency, according to SSH analysis,
were blotted in duplicate to nylon membranes. One set of Se-adequate and
Se-deficient blotted cDNAs was probed with 32P-labeled Se-adequate sub-
tracted cDNA. The duplicate set was probed with 32P-labeled Se-deficient
subtracted cDNA. Circles indicate two species whose expression is clearly en-
hanced in Se-deficient samples.
between the two samples in the number of cycles required to reach the
threshold value. The regression equations for the exponential curves
that fit each data set were used to calculate the number of cycles
required to reach the threshold value. Figure 2 shows one example
of this analysis. Repeated experiments showed that, on average, Se-
deficient samples required 2.86 more cycles to reach the detection
threshold than did Se-adequate samples. This corresponds to an aver-
age difference in abundance of 7.3-fold (22.86). This estimate of differ-
ence in abundance is in good agreement with previously published
results from our laboratory derived from Northern blots showing that
mRNA for GPX1 in Se-deficient rat liver is only 5–15% as abundant as
it is in Se-adequate liver (28,33).
Figure 3 shows similar graphs for Tf, transferrin receptor (TfR),
and iron-regulatory protein (IRP) 1. Each of these graphs summarizes
Biological Trace Element Research
Vol. 74, 2000

64
Christensen et al.
Fig. 2. Relative multiplex RT-PCR analysis of glutathione peroxidase
(GPX1) gene expression in Se-adequate and Se-deficient rat liver. Total RNA
from Se-adequate and Se-deficient rat livers was reverse transcribed followed by
PCR amplification using primers for GPX1 and 18S rRNA. The intensity of GPX1
signal was normalized to the intensity of signal for 18S rRNA and the normal-
ized signal was plotted as a function of cycle number. The −Se sample took
nearly three cycles more to reach the threshold value, indicating a much lower
abundance of GPX1 mRNA.
data from three or more experiments on the same samples. For each
gene, the −Se sample reached the threshold value before the +Se sam-
ple. This represents a higher level in Se deficiency than in Se adequacy
of mRNA in rat liver for each of these proteins involved in Fe metabo-
lism. For Tf, mRNA levels averaged 30% higher in Se deficiency than
in Se adequacy; for TfR, it was 32%; and for IRP 1, it was 63%. No
consistent differences were shown for ferritin light or heavy chain, or
for IRP 2.
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
65
Biological Trace Element Research
Vol. 74, 2000

66
Christensen et al.
DISCUSSION
Interactions between Fe and Se have been explored in previous stud-
ies. High intakes of Fe by rats have little effect on liver GPX1 in spite of
increased lipid peroxidation (34). Likewise, high Fe intakes do not ad-
versely alter tissue distribution or concentrations of Se (35). In contrast,
deficiency of Fe appears to decrease tissue Se concentration and GPX1
activity (36,37). Dietary deficiency of Se does not significantly affect mea-
sures of Fe status (38). Morphological changes in red blood cells have
been reported (39), but such changes did not result in anemia. In one
study, chronic Se deficiency resulted in reduced Fe-binding capacity of
rat plasma (40). Our findings that hematocrit and hemoglobin were unaf-
fected by Se intake are consistent with previous reports.
Iron induces lipid peroxidation in animal and cell-culture models of
oxidative stress (for recent examples, see refs. 41–43). In the Fenton reac-
tion (H2O2 + Fe2+ → OH− + OH• + Fe3+), the highly reactive OH radical
is produced, which nonspecifically targets DNA, lipids, and proteins.
Members of the GPX family assist in antioxidant defense by reducing
hydrogen peroxide and lipid hydroperoxides to their corresponding
alcohols, using glutathione as a reducing agent (R–OOH + 2GSH →
R–OH + GSSG + H2O). Diets deficient in Se reduce activity of GPX fam-
ily members, thus reducing the capacity to neutralize hydrogen peroxide.
An increase in hydrogen peroxide would promote formation of Fenton
reaction products. Thus, Se deficiency may increase oxidative stress by
allowing accumulation or production of reactive oxygen species.
Cytosolic Fe concentrations reflect the balance between cellular
uptake of iron, by means of TfRs, and sequestration of iron in the form
of ferritin. The synthesis of ferritin and of TfRs is regulated posttran-
scriptionally by iron. When cytosolic Fe levels fall, IRPs bind to iron-
responsive elements (IREs) in the 5Ј and 3Ј untranslated regions (UTRs)
of ferritin and TfR mRNAs, respectively (44,45). Binding of IRPs to the
IRE in the 5Ј UTR of ferritin inhibits its translation, whereas binding to
the IRE in the 3Ј UTR of TfR stabilizes its mRNA. The net effect of these
changes is to increase cytosolic Fe availability by increasing its uptake
from extracellular sources and reducing its sequestration within the cell.
Two IRPs have been identified. IRP 1 functions alternatively as an
RNA binding protein and as the cytosolic isoform of the mitochondrial
Krebs cycle enzyme aconitase (46,47). IRP 2 has recently been shown to
be sufficient to regulate expression of ferritin and TfR in cells lacking
IRP 1 (48). Phosphorylation of IRPs increases their affinity for IRE mRNA
binding (49,50). Nitric oxide also increases IRE binding by IRP 1, but not
by IRP 2 (51).
The complex and intricate interactions of IRPs with IREs in response
to various stimuli provide a means of limiting reactive free iron in cells.
The logical response of the cell to reactive oxygen species would be to
sequester iron to minimize the risk of oxidative stress (52). In vitro studies
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
67
have demonstrated an inhibition of IRP activity by xanthine oxidase-
•−
stimulated production of O2 and H2O2 (53,54). However, the response
to reactive oxygen species seen in cells is quite different from that ob-
served in cell-free assays of lysates or extracts.
Exposure of cells to H2O2 results in a rapid activation of IRP 1 not
seen in cell extracts (55,56). This in vivo activation is unaffected by iron
chelators or hydroxyl radical scavengers, but it is blocked by inhibitors
of protein phosphatases. Pantopoulos et al. (57) have demonstrated dif-
ferential regulation of IRP 1 by intracellular and extracellular H2O2. In
their studies, extracellular H2O2 rapidly activated IRP 1 without increas-
ing intracellular H2O2. Inhibition of catalase increased intracellular H2O2
but did not increase IRP 1 activation. In more recent experiments (58), the
same investigators reconstituted the response of IRP 1 to extracellular
H2O2 in an in vitro system. These experiments showed that IRP 1 activa-
tion required the presence of a nonsoluble component in the cell pellet.
Their results, combined with the data on protein phosphatase-mediated
activation, suggest that H2O2 must be sensed to initiate or stimulate a
pathway leading to IRP 1 activation.
Results from the present study are consistent with the previous
demonstration of H2O2 activation of IRP 1 in cells. In our whole-animal
model, Se deficiency virtually abolished liver GPX1 activity. The effect of
Se deficiency on expression of genes for IRP 1 and TfR is consistent with
an elevation of intracellular H2O2, which would be expected when GPX1
activity is reduced. In the work of Pantopoulos et al. (57), extracellular
H2O2 had a more pronounced effect on IRP 1 activity than did intracel-
lular H2O2. Our previous work showed that activity of plasma GPX
(GPX3) in rats was reduced in Se deficiency to only 3% of the activity
found in Se adequacy. Thus, extracellular H2O2 would also be expected
to be elevated under dietary conditions employed in this study.
Expression of genes for ferritin subunits is modified by factors in
addition to IRPs. Simultaneous administration of thyroid hormone (T3)
augments the iron-induced reduction of IRP binding to ferritin mRNA
(59). Coordinate and opposite control of ferritin and TfR by IRPs is not
seen in rat liver regeneration (60). In addition, IRPs have differential
effects on abundances of transcripts containing IREs (61,62). Thus,
although our observation of a lack of coordinate regulation of TfR and
ferritin gene expression by Se does not fit the standard paradigm, it is
not without precedent.
In summary, dietary Se deficiency resulted in modest but consistent
increases in expression of genes for Tf, TfR and IRP 1 in rat liver. These
increases occurred in the absence of significant effects on indicators of Fe
status. Deficiency of Se results in dramatically reduced activity of GPX. As
a result, H2O2 and other reactive oxygen species accumulate. Increased
H2O2 activates IRP 1 binding to IREs in mRNAs for TfR and ferritin, typ-
ically increasing and decreasing their synthesis, respectively. The lack of
effect on ferritin gene expression seen in this work is consistent with other
Biological Trace Element Research
Vol. 74, 2000

68
Christensen et al.
studies in which the standard paradigm of coordinate but opposite regu-
lation of TfR and ferritin did not apply. These findings suggest a possible
role for dietary Se in moderating iron metabolism.
REFERENCES
1. S. C. Low and M. J. Berry, Knowing when not to stop: selenocysteine incorporation
in eukaryotes, Trends. Biochem. Sci. 21, 203–208 (1996).
2. J. T. Rotruck, A. L. Pope, H. E. Ganther, A. B. Swanson, D. G. Hafeman, and W. G.
Hoekstra, Selenium: biochemical role as a component of glutathione peroxidase, Sci-
ence 179, 588–590 (1973).
3. F. F. Chu, J. H. Doroshow, and R. S. Esworthy, Expression, characterization, and tis-
sue distribution of a new cellular selenium-dependent glutathione peroxidase,
GSHPx–GI., J. Biol. Chem. 268, 2571–2576 (1993).
4. K. R. Maddipati, C. Gasparski, and L. J. Marnett, Characterization of the hydroper-
oxide-reducing activity of human plasma, Arch. Biochem. Biophys. 254, 9–17 (1987).
5. F. Ursini, M. Maiorino, M. Valente, L. Ferri, and C. Gregolin, Purification from pig
liver of a protein which protects liposomes and biomembranes from peroxidative
degradation and exhibits glutathione peroxidase activity on phosphatidylcholine
hydroperoxides, Biochim. Biophys. Acta 710, 197–211 (1982).
6. L. Hall, K.Williams, A. C. F. Perry, J. Frayne, and J. A. Jury, The majority of human
glutathione peroxidase type 5 (GPX5) transcripts are incorrectly spliced: implications
for the role of GPX5 in the male reproductive tract, Biochem. J. 333, 5–9 (1998).
7. M. J. Berry, L. Banu, and P. R. Larsen, Type I iodothyronine deiodinase is a seleno-
cysteine-containing enzyme, Nature 349, 438–440 (1991).
8. J. C. Davey, K. B. Becker, and M. J. Schneider, Cloning of a cDNA for the type II
iodothyronine deiodinase, J. Biol. Chem. 270, 26,786–26,789 (1995).
9. W. Croteau, S. L. Whittemore, M. J. Schneider, and D. L. St. Germain, Cloning and
expression of a cDNA for a mammalian type III iodothyronine deiodinase, J. Biol.
Chem. 270, 16,569–16,575 (1995).
10. P. Y. Gasdaska, J. R. Gasdaska, S. Cochran, and G. Powis, Cloning and sequencing of
a human thioredoxin reductase, FEBS Lett. 373, 5–9 (1995).
11. M. A. Correia and R. F. Burk, Defective utilization of haem in selenium-deficient rat
liver, Biochem. J. 214, 53–58 (1983).
12. K. E. Hill and R. F. Burk, Effect of selenium deficiency and vitamin E deficiency on
glutathione metabolism in isolated rat hepatocytes, J. Biol. Chem. 257, 10,668–10,672
(1982).
13. R. F. Burk, A. M. MacKinnon, and F. R. Simon, Selenium and hepatic microsomal
hemoproteins, Biochem. Biophys. Res. Commun. 56, 431–436 (1974).
14. R. Reiter and A. Wendel, Selenium and drug metabolism. I. Multiple modulations of
mouse liver enzymes, Biochem. Pharmacol. 32, 3063–3067 (1983).
15. R. A. Goyer, Toxic and essential metal interactions, Annu. Rev. Nutr. 17, 37–50 (1997).
16. R. C. McKenzie, T. S. Rafferty, and G. J. Beckett, Selenium: an essential element for
immune function, Immunol. Today 19, 342–345 (1998).
17. G. F. Combs, Jr. and W. P. Gray, Chemopreventive agents: selenium, Pharmacol. Ther.
79, 179–192 (1998).
18. I. Vietor and L. A. Huber, In search of differentially expressed genes and proteins.
Biochim. Biophys. Acta 1359, 187–199 (1997).
19. C. G. Sagerstrom, B. I., Sun and H. L. Sive, Subtractive cloning: past, present, and
future. Annu. Rev. Biochem. 66, 751–783 (1997).
20. P. Liang and A. B. Pardee, Differential display of eukaryotic messenger RNA by
means of the polymerase chain reaction, Science 257, 967–971 (1992).
21. J. Welsh, K. Chada, S. S. Dalal, R. Cheng, D. Ralph, and M. McClelland, Arbitrarily
primed PCR fingerprinting of RNA, Nucl. Acids Res. 20, 4965–4970 (1992).
Biological Trace Element Research
Vol. 74, 2000

Selenium and Gene Expression
69
22. L. Diatchenko, Y. F. C. Lau, A. P. Campbell, A. Chenchik, F. Moqadam, B. Huang,
et al., Suppression subtractive hybridization: a method for generating differentially
regulated or tissue-specific cDNA probes and libraries, Proc. Natl. Acad. Sci. USA
93, 6025–6030 (1996).
23. National Research Council, Guide for the Care and Use of Laboratory Animals, National
Academy Press, Washington, DC, p. 125 (1996).
24. P. G. Reeves, F. H. Nielsen, and G. C. Fahey Jr., AIN-93 purified diets for laboratory
rodents: final report of the American Institute of Nutrition Ad Hoc Writing Commit-
tee on the Reformulation of the AIN-76A Rodent Diet, J. Nutr. 123, 1939–1951 (1993).
25. National Research Council, Nutrient Requirements of Laboratory Animals, National
Academy Press, Washington, DC, p. 173 (1995).
26. Expert Scientific Working Group, Summary of a report on assessment of iron nutri-
tional status of the United States population, Am. J. Clin. Nutr. 42, 1318–1330 (1985).
27. J. D. Torrance and T. H. Bothwell, A simple technique for measuring storage iron con-
centrations in formalinised liver samples, S. Afr. J. Med. Sci. 33, 9–11 (1968).
28. M. J. Christensen, P. Cammack, and C. Wray, Tissue specificity of selenoprotein gene
expression in rats, J. Nutr. Biochem. 6, 367–372 (1995).
29. R. A. Lawrence and R. F. Burk, Glutathione peroxidase activity in selenium-deficient
rat liver, Biochem. Biophys. Res. Commun. 71, 952–958 (1976).
30. M. J. Christensen, B. L. Nelson, and C. D. Wray, Regulation of glutathione S-transferase
gene expression and activity by dietary selenium, Biochem. Biophys. Res. Commun. 202,
271–277 (1994).
31. W. H. Habig, M. J. Pabst, and W. B. Jakoby, Glutathione S-transferases. The first enzy-
matic step in mercapturic acid formation, J. Biol. Chem. 249, 7130–7139 (1974).
32. O. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, Protein measurement
with the folin phenol reagent, J. Biol. Chem. 193, 265–275 (1951).
33. M. J. Christensen and K. W. Burgener, Dietary selenium stabilizes glutathione perox-
idase mRNA in rat liver, J. Nutr. 122, 1620–1626 (1992).
34. Y. H. Lee, D. K. Layman, R. R. Bell, and H. W. Norton, Response of glutathione per-
oxidase and catalase to excess dietary iron in rats, J. Nutr. 111, 2195–2202 (1981).
35. A. G. Abdel Rahim, J. R. Arthur, and C. F. Mills, Effects of dietary copper, cadmium,
iron, molybdenum and manganese on selenium utilization by the rat, J. Nutr. 116,
403–411 (1986).
36. Y. H. Lee, D. K. Layman, and R. R. Bell, Glutathione peroxidase activity in iron-
deficient rats, J. Nutr. 111, 194–200 (1981).
37. P. M. Moriarty, M. F. Picciano, J. L. Beard, and C. C. Reddy, Classical selenium-
dependent glutathione peroxidase expression is reduced secondary to iron deficiency
in rats, J. Nutr. 125, 293–301 (1995).
38. N. Chareonpong-Kawamoto and K. Yasumoto, Iron-induced changes in electrocar-
diogram in selenium-deficient rats, Nutr. Res. 16, 961–968 (1996).
39. R. Boyne and J. R. Arthur, Anaemia and changes in erythrocyte morphology associ-
ated with copper and selenium deficiencies and dietary restriction in rats, Res. Vet.
Sci. 49, 151–156 (1990).
40. M. G. Doni, A. Falanga, F. Delaini, E. Vicenzi, M. Tomasiak, and M. D. Donati, The
effect of vitamin E or selenium on the oxidant-antioxidant balance in rats, Br. J. Exp.
Pathol. 65, 75–80 (1984).
41. I. Sziraki, K. P. Mohanakumar, P. Rauhala, H. G. Kim, K. J. Yeh, and C. C. Chiueh,
Manganese: a transition metal protects nigrostriatal neurons from oxidative stress in
the iron-induced animal model of parkinsonism, Neuroscience 85, 1101–1111 (1998).
42. Y. Ma, D. Zhang, I. Kawabata, T. Kiriu, S. Toyokuni, K. Uchida, et al., Copper and
iron-induced oxidative damage in non-tumor bearing LEC rats, Pathol. Int. 47,
203–208 (1997).
43. O. Sergent, B. Griffon, I. Morel, M. Chevanne, M. P. Dubos, P. Cillard, et al., Effect of
nitric oxide on iron-mediated oxidative stress in primary rat hepatocyte culture,
Hepatology 25, 122–127 (1997).
44. T. Rouault and R. Klausner, Regulation of iron metabolism in eukaryotes, Curr. Top.
Cell. Regul. 35, 1–19 (1997).
Biological Trace Element Research
Vol. 74, 2000

70
Christensen et al.
45. J. L. Beard, H. Dawson, and D. J. Piñero, Iron metabolism: a comprehensive review,
Nutr. Rev. 54, 295–317 (1996).
46. B. R. Henderson, Iron regulatory proteins 1 and 2, Bioessays 18, 739–746 (1996).
47. M. C. Kennedy, L. Mende-Meuller, G. A. Blondin, and H. Beinert, Purification and
characterization of cytosolic aconitase from beef liver and its relationship to the iron-
responsive element binding protein, Proc. Natl. Acad. Sci. USA 89, 11,730–11,734
(1992).
48. K. L. Schalinske, K. P. Blemings, D. W. Steffen, O. S. Chen, and R. S. Eisenstein, Iron
regulatory protein 1 is not required for the modulation of ferritin and transferrin
receptor expression by iron in a murine pro-B lymphocyte cell line, Proc. Natl. Acad.
Sci. USA 94, 10,681–10,686 (1997).
49. R. S. Eisenstein, P. T. Tuazon, K. L. Schalinske, S. A. Anderson, and J. A. Traugh, Iron-
responsive element-binding protein. Phosphorylation by protein kinase C., J. Biol.
Chem. 268, 27,363–27,370 (1993).
50. K. L. Schalinske and R. S. Eisenstein, Phosphorylation and activation of both iron reg-
ulatory proteins 1 and 2 in HL-60 cells, J. Biol. Chem. 271, 7168–7176 (1996).
51. J. D. Phillips, D. V. Kinikini, Y. Yu, B. Guo, and E. A. Leibold, Differential regulation
of IRP1 and IRP2 by nitric oxide in rat hepatoma cells, Blood 87, 2983–2992 (1996).
52. T. A. Rouault and R. D. Klausner, The impact of oxidative stress on eukaryotic iron
metabolism, EXS 77, 183–197 (1996).
53. G. Cairo, E. Castrusini, G. Minotti, and A. Bernelli-Zazzer, Superoxide and hydrogen
peroxide-dependent inhibition of iron regulatory protein activity: a protective strata-
gem against oxidative injury, FASEB J. 10, 1326–1335 (1996).
54. G. Cairo, L. Tacchini, S. Recalcati, B. Azzimonti, G. Minotti, and A. Bernelli-Zazzer,
Effect of reactive oxygen species on iron regulatory protein activity, Ann. NY Acad.
Sci. 851, 179–186 (1998).
55. K. Pantopoulos and M. W. Hentze, Rapid responses to oxidative stress mediated by
iron regulatory protein, EMBO J. 14, 2917–2924 (1995).
56. E. A. L. Martins, R. L. Robalinho, and R. Meneghini, Oxidative stress induces acti-
vation of a cytosolic protein responsible for control of iron uptake, Arch. Biochem. Bio-
phys. 316, 128–134 (1995).
57. K. Pantopoulos, S. Mueller, A. Atzberger, W. Ansorge, W. Stremmel, and M. W.
Hentze, Differences in the regulation of iron regulatory protein-1 (IRP-1) by extra-
and intracellular oxidative stress, J. Biol. Chem. 272, 9802–9808 (1997).
58. K. Pantopoulos and M. W. Hentze, Activation of iron regulatory protein-1 by oxida-
tive stress in vitro, Proc. Natl. Acad. Sci. USA 95, 10,559–10,563 (1998).
59. P. J. Leedman, A. R. Stein, W. W. Chin, and J. T. Rogers, Thyroid hormone modu-
lates the interaction between iron regulatory proteins and the ferritin mRNA iron-
responsive element, J. Biol. Chem. 271, 12,017–12,023 (1996).
60. G. Cairo, L. Tacchini, and A. Pietrangelo, Lack of coordinate control of ferritin and
transferrin receptor expression during rat liver regeneration, Hepatology 28, 173–178
(1998).
61. O. S. Chen, K. L. Schalinske, and E. S. Eisenstein, Dietary iron modulates the activ-
ity of iron regulatory proteins and the abundance of ferritin and mitochondrial aconi-
tase in rat liver, J. Nutr. 127, 238–248 (1997).
62. K. L. Schalinske, O. S. Chen, and R. S. Eisenstein, Iron differentially stimulates trans-
lation of mitochondrial aconitase and ferritin mRNAs in mammalian cells. Implica-
tions for iron regulatory proteins as regulators of mitochondrial citrate utilization, J.
Biol. Chem. 273, 3740–3746 (1998).
Biological Trace Element Research
Vol. 74, 2000